Asymmetries in behavioral and neural responsesto spectral cues demonstrate the generalityof auditory looming biasRobert Baumgartnera,b,1, Darrin K. Reedc, Brigitta Tóthd, Virginia Besta,e, Piotr Majdakb, H. Steven Colburna,c,and Barbara Shinn-Cunninghama,c

aHearing Research Center, Boston University, Boston, MA 02215; bAcoustics Research Institute, Austrian Academy of Sciences, Vienna 1040, Austria;cDepartment of Biomedical Engineering, Boston University, Boston, MA 02215; dInstitute of Cognitive Neuroscience and Psychology, Hungarian Academyof Sciences, Budapest 1117, Hungary; and eDepartment of Speech, Language & Hearing Sciences, Boston University, Boston, MA 02215

Edited by Dale Purves, Duke University, Durham, NC, and approved July 28, 2017 (received for review February 25, 2017)

Studies of auditory looming bias have shown that sources in-creasing in intensity are more salient than sources decreasing inintensity. Researchers have argued that listeners are more sensitiveto approaching sounds compared with receding sounds, reflectingan evolutionary pressure. However, these studies only manipulatedoverall sound intensity; therefore, it is unclear whether looming biasis truly a perceptual bias for changes in source distance, or only insound intensity. Here we demonstrate both behavioral and neuralcorrelates of looming bias without manipulating overall soundintensity. In natural environments, the pinnae induce spectral cuesthat give rise to a sense of externalization; when spectral cuesare unnatural, sounds are perceived as closer to the listener. Wemanipulated the contrast of individually tailored spectral cues tocreate sounds of similar intensity but different naturalness. Weconfirmed that sounds were perceived as approaching when spectralcontrast decreased, and perceived as recedingwhen spectral contrastincreased. Wemeasured behavior and electroencephalographywhilelisteners judged motion direction. Behavioral responses showeda looming bias in that responses were more consistent for soundsperceived as approaching than for sounds perceived as receding. Ina control experiment, looming bias disappeared when spectralcontrast changes were discontinuous, suggesting that perceivedmotion in distance and not distance itself was driving the bias.Neurally, looming bias was reflected in an asymmetry of late event-related potentials associated with motion evaluation. Hence, bothour behavioral and neural findings support a generalization of theauditory looming bias, representing a perceptual preference forapproaching auditory objects.

auditory looming bias | electroencephalography | distance motionperception | sound externalization | head-related transfer functions

Imagine yourself alone in the wilderness. Suddenly, a threat-ening sound permeates the darkness. Is it approaching? This is

a critical question when it comes to your survival because appro-aching objects usually pose a greater threat than receding objects(1). The phenomenon that approaching sounds are more salientthan receding sounds is commonly termed “auditory looming bi-as.” Looming bias is reflected in a broad variety of psychophysicaltasks related to salience and alertness: bias in loudness-changeestimates (2–4) and judgments of duration (5), improved discrim-inability of motion speed (6), underestimated distances for egocen-trically moving (4) or bypassing sounds (7, 8), and reduced reactiontime for auditory (3, 9) and visual (3) targets preceded by loomingsounds. In animals, looming biases result in faster learning speedduring associative conditioning (10) and longer duration of attention(11). This list shows that looming bias triggers a variety of perceptsacross a wide range of psychoacoustic tasks. Despite its broad be-havioral significance, the mechanisms underlying auditory loomingbias are still poorly understood.A universal problem of previous research on auditory looming

bias is that source distance was manipulated using overall sound

intensity (sounds increasing in intensity perceived as approach-ing). Hence, it is unclear whether looming bias actually reflects abias in the perception of intensity changes or distance motion.Moreover, it is hard to disentangle the contribution of simpleneural nonlinearities in response to intensity gradients (12, 13)from higher-level perceptual asymmetries of motion in distance.Previous studies addressed this issue by comparing sounds ofdifferent spectral structure: complex tones elicited stronger be-havioral looming bias (2, 11) and larger neural differences (14–16) than did noise with equal intensity increase, indicating thatintensity changes did not directly cause looming bias. However,studies with a different spatial task (17) or species (10) found theopposite effect (i.e., stronger looming bias for noise than tonalsounds), calling into question these results.An early fMRI study that contrasted static sounds with sounds

increasing or decreasing in intensity found activity in the rightplanum temporale (4), which is associated with processing of au-ditory motion in any spatial direction (18, 19), consistent with theidea that sound intensity changes can be perceived as motion indistance. The contrast between sounds increasing vs. decreasing inintensity, however, revealed a different network, including supe-rior temporal sulcus and amygdala, which may reflect a warningprocess for approaching objects (4, 9, 20). A more recent fMRI

Significance

Previous studies demonstrated “auditory looming bias” exclu-sively by manipulating overall sound intensity. Hence, it is notclear whether this bias truly reflects perceptual differences insensitivity to motion direction rather than changes in intensity.We manipulated individualized spectral cues to create stimulithat were perceived as either approaching or receding, whilecontrolling loudness. We assessed discrimination of motiondirection and analyzed simultaneously recorded neural re-sponses using electroencephalography. Our results show bothbehavioral and neural evidence of looming bias. Therefore, ourstudy demonstrates that the bias is truly about perceived mo-tion in distance, not intensity changes.

Author contributions: R.B., D.K.R., B.T., V.B., P.M., H.S.C., and B.S.-C. designed research;R.B. performed research; R.B. analyzed data; and R.B., D.K.R., B.T., V.B., P.M., H.S.C., andB.S.-C. wrote the paper.

The authors declare no conflict of interest.

This article is a PNAS Direct Submission.

Freely available online through the PNAS open access option.

Data deposition: Code for stimulus creation (sig_baumgartner2017looming), HRTFs, andexperimental results (both data_baumgartner2017looming) are integrated in the Audi-tory Modeling Toolbox (amtoolbox.sourceforge.net). EEG recordings and analysis scriptsare provided via Zenodo (https://dx.doi.org/10.5281/zenodo.832899).1To whom correspondence should be addressed. Email: robert.baumgartner@oeaw.ac.at.

This article contains supporting information online at www.pnas.org/lookup/suppl/doi:10.1073/pnas.1703247114/-/DCSupplemental.

www.pnas.org/cgi/doi/10.1073/pnas.1703247114 PNAS | September 5, 2017 | vol. 114 | no. 36 | 9743–9748

PSYC

HOLO

GICALAND

COGNITIVESC

IENCE

S

Dow

nloa

ded

by g

uest

on

Nov

embe

r 8,

202

0

study (21) and neural recordings in animals (10, 14) reported dif-ferences related to looming bias in low-level auditory corticalareas. Whether these differences are caused by lateral or top-down processes is unclear. The processing latencies reported inprevious studies appear to be inconsistent: some found multisen-sory enhancements in visual cortical areas at early sensory levels(80-ms postsound onset) (15, 22), whereas others found no sig-nificant differences in global field strength before 600 ms (16). Amore instantaneous change of distance cues (rather than gradualchanges of sound intensity) might be better suited to identify thetime course of auditory looming bias.Auditory event-related potential (ERP) studies of horizontal

and vertical motion perception support a stereotypical “motiononset response” that consists of two consecutive stages associ-ated with different functions (23–26): a negative deflection oc-curring roughly 100 ms after the evoking event, denoted by N1,represents an early, presumably automatic, sensory detectionof spatial changes, and a positive deflection occurring roughly200 ms after the evoking event, denoted by P2, likely representsevaluation of motion direction. Looming bias arguably relies onthe evaluation of motion direction. Hence, if the motion onsetresponse can be generalized to motion in distance, then loomingbias may be reflected in P2.Here, we simultaneously measured behavior and neural re-

sponses to explore whether (i) changes in overall sound intensityand (ii) stimulus continuity are required to activate looming bias,and (iii) whether auditory looming bias is linked to early sensoryor the later stages of neural processing. Based on the hypothesisthat looming bias reflects a gain in perceptual detectability ofapproaching objects relative to receding objects, we measuredlooming bias in a motion direction discrimination task. Hence, weaimed to assess the origin of looming bias more directly thanprevious studies that measured consequential effects of the biasrelated to stimulus salience and alertness. We used spectralchanges to create sounds perceived as approaching or recedingbetween external, internal, and intermediate positions (coloredred, blue, and green, respectively, in Fig. 1 A and B) by reducingthe contrast of measured high-frequency spectral cues from thenatural acoustics of each individual listener. Spectral cues areparticularly suitable for this investigation because (i) in contrast togradual intensity changes, spectral cues can evoke instantaneouschanges of distance percepts (27–30) and thus allow preciseanalysis of processing latencies; (ii) spectral cue manipulations cancreate transitions to internal (infinitely close) auditory perceptswithin peripersonal space where looming bias is most prominent(31); and (iii) spectral cues can be manipulated independently ofoverall stimulus intensity.

ResultsLoudness Predictions. We used predictions of a loudness model(32) to assure that loudness changes cannot explain the loomingbias caused by spectral contrast switches. Fig. 1B shows the effectof our contrast manipulation technique on the stimulus magnitudespectrum (Top) and on the corresponding predictions of relativeloudness changes (Bottom) for all stimuli of experiment (Exp.)I (individual data provided in Interindividual Comparison ofFrequency-Specific Loudness Changes and Fig. S1). Flattening thereference magnitude spectrum (C = 1, red) decreased frequency-specific loudness in the lower two to three octaves of the stimulus(1–6 kHz) and increased it only in the highest octave (8–16 kHz).On average across frequencies (Lower Right), flattening the ref-erence magnitude spectrum decreased predicted overall loudness.Hence, changes in loudness perception, if they occurred, would beexpected to oppose the typical loudness-induced distance percept.Interaural level differences (ILDs) are also known to act as cues

for distance perception (18, 29). Relative changes in frequency-specific ILDs can be easily extracted from loudness predictions(Fig. 1B, Bottom) as the difference between ipsilateral (solid lines)

and contralateral (dashed lines) loudness. The predictions showthat a decrease in spectral contrast was accompanied by an ILDincrease at the lower two octaves and an ILD decrease at thehigher two octaves.

Behavioral Results. In two experiments, listeners judged switchesbetween spectral contrasts (C1 and C2) in forced choice paradigms(Fig. 1C). Experiment I allowed subjects three response alterna-tives [3-alternative forced-choice (AFC): “approaching,” “reced-ing,” or “static”] and tested instantaneous spectral contrast switches

A

B

C

Fig. 1. Contrast of spectral cues naturally induced by the pinnae and torsowas manipulated to create approaching and receding sounds. (A) Spatialconfiguration of hypothesized spatial percepts with number of listenerstested per source angle (left, front, and right) in Exps. I and II. (B) Effect ofspectral contrast manipulation according to factor C on magnitude re-sponses of listener-specific stimuli of Exp. I (B, Top) as well as their frequency-specific (B, Bottom Left) and overall (B, Bottom Right) loudness changesrelative to C = 1. Shaded areas denote ±1 SEM (n = 15). Note that changes inoverall loudness oppose the intended effect of contrast switch. (C) Schematicrepresentation of an experimental trial in Exp. I and Exp. II: starting withnoise filtered according to C1, then cross-faded to C2, followed by an openresponse period, and a jittered intertrial interval. The left–right arrow in Exp.I represents a temporal jitter of ±50 ms. Listeners were instructed to reportwhether the sound was perceived as approaching, receding, or static in Exp. Iand only approaching or receding in Exp. II.

9744 | www.pnas.org/cgi/doi/10.1073/pnas.1703247114 Baumgartner et al.

Dow

nloa

ded

by g

uest

on

Nov

embe

r 8,

202

0

[using an interstimulus interval (ISI) of zero] while neural re-sponses were measured using electroencephalography (EEG).Experiment II tested speeded responses for two alternatives(2-AFC: approaching or receding) and presented trials with eitherinstantaneous but continuous (ISI of 0 ms) or discontinuous (ISI of100 ms) changes in spectral shape across stimulus pairs. ExperimentII served as a control to test whether looming bias is truly linked tothe perception of sounds moving in distance rather than the merepercept of source proximity. Some listeners perceived spectralcontrast manipulations as elevation changes at some source angles.Therefore, for each individual, we selected the source angle that ledto the most consistent percept of a change in distance, based on ashort initial run of the discrimination task (pretest) using threesource angles (left, front, and right). Fig. 1A shows the source anglesthat were selected for the subsequent looming experiments.Although listeners never received feedback about the hypoth-

esized distance change, decreasing spectral contrasts (C1 > C2,solid orange lines in Fig. 2) were more likely perceived asapproaching (filled orange triangles) and increasing spectralcontrasts (C1 < C2, dashed green lines) as receding (open greentriangles). These dominant associations were used for subsequentstatistical analyses and were also observed in the pretest results(Behavioral Results of Pretest for Individual Source Angle Selectionand Fig. S2). Trials without spectral changes (C1 = C2) in Exp. Iwere almost always perceived as “static,” confirming that the lis-teners remained vigilant and performed the task reliably.In Exp. I (Fig. 2, Left), response consistency (percentage of

responses for the same percept category), used here as a measureof cue salience, differed significantly between contrast pairs,i.e., the combination of spectral contrasts independent of pre-sentation order (0↔ 1 vs. 0↔ 0.5 vs. 0.5↔ 1) (F1.4,20 = 14, P < 0.01,η2p = 0.51). Listeners were most consistent for the largest contrastswitches (C: 0 ↔ 1) and least consistent for pairs consisting of the

two higher contrast values (C: 0.5 ↔ 1; Fig. 2 shows significancelevels of paired comparisons). Importantly, response consistencyalso differed significantly between switch directions, i.e., the pre-sentation order within contrast pairs (C1<C2 vs. C1>C2) (F1,14= 6.9,P = 0.020, η2p = 0.33). Listeners were more consistent in la-beling decreasing spectral contrasts than increasing spectral contrasts;this pattern is consistent with looming bias in that approachingsounds elicited more consistent identification responses than didreceding sounds.Response consistency in Exp. II also differed significantly

between contrast pairs (F1.9,22 = 40, P < 0.001, η2p = 0.77). As inExp. I, judgments were least consistent for pairs consisting of thetwo higher contrast values (C: 0.5 ↔ 1). Overall response con-sistency was not affected by stimulus continuity (ISI = 0 ms vs.ISI = 100 ms) (F1,12 = 0.55, P = 0.47, η2p = 0.044), but therewere tendencies of a main effect of switch direction (F1,12 = 4.7,P = 0.052, η2p = 0.28) and an interaction between stimuluscontinuity and switch direction (F1,12 = 3.9, P = 0.072, η2p = 0.25).While interindividual differences in the amount of looming biaswere generally large [continuity (cont.): M = 15%, SD = 19%;discontinuity (discont.): M = −0.32%, SD = 19%], only 1 of 13listeners showed a markedly larger looming bias for discontinu-ous stimuli compared with continuous stimuli (data provided inOutlier Evaluation for Experiment II and Fig. S3). The interactionbecame highly significant after removing this one outlier fromthe analysis (F1,12 = 15, P < 0.01, η2p = 0.57). Removal of thisoutlier had no significant effect on the other statistics.Response times were measured with respect to the onset of

spectral contrast switches and were evaluated only for the dominantresponse associations (data provided in Response Times in Experi-ments I and II and Fig. S4). In Exp. I, early response times (25%percentiles) were slightly faster for decreasing compared with in-creasing contrasts (F1,14 = 5.2, P = 0.039, η2p = 0.27) and for largercontrast switches (F1.5,21 = 6.0, P = 0.014, η2p = 0.30). Differences inresponse times for decreasing vs. increasing contrasts were smallerfor later responses (30–90% percentiles). In Exp. II with speededresponses, variance in response times dramatically increased andneither early nor late response times differed significantly.

Event-Related Potentials. We investigated the neural underpin-nings of auditory looming bias by means of ERPs. Fig. 3 showsthe ERPs elicited by stimulus onset and spectral switch. Thestimulus onset (Fig. 3A) was accompanied by a frontocentralnegativity (onset-N1) followed by a central positivity (onset-P2).Average onset-N1 and onset-P2 amplitudes were measured foreach spectral contrast (C1) at the Cz site within the intervals of80–140 ms and 140–280 ms, respectively, as determined by zero-crossings of the grand average. Onset-N1 amplitudes were not sig-nificantly different across spectral contrasts (F1.6,23 = 1.9, P = 0.18,η2p = 0.12). Onset-P2 amplitudes were significantly larger for theoriginal contrast (C1 = 1) compared with the flattened contrast(C1 = 0) but not to the intermediate contrast (C1 = 0.5; see Fig. 3for significance levels) (F1.9,26 = 6.7, P < 0.01, η2p = 0.32).The latencies of ERP components elicited by switches in

spectral contrast were similar to the stimulus onset and werechosen at 90–150 ms for switch-N1 and 150–290 ms for switch-P2(Fig. 3B). As for the onset-ERPs, these time ranges were deter-mined on the basis of zero-crossings of the grand average; theonly exception was the P2 end point, which had to be extrapolatedbecause no final zero-crossing occurred. Switch trials (C1 ≠ C2) allelicited N1–P2 complexes. Constant trials with no contrast switch(C1 = C2) elicited neither N1 nor P2 activity (dash-dotted line inFig. 3B, Lower Left); these trials were therefore excluded fromstatistical analyses. Switch-N1 amplitudes differed significantlyacross contrast pairs (F1.9,26 = 4.6, P = 0.022, η2p = 0.25) butnot switch direction (F1,14 = 1.8, P = 0.20, η2p = 0.12). After multiplecomparison correction of post hoc paired comparisons, none of theswitch-N1 amplitudes were significantly different across contrast

ResponseSpectral contrast

decrease increase

staticSpectral contrast pair

Exp. II: continuous Exp. II: discontinuous

* n.s.

Spectral contrast pair

******

*********

**

*

"Behaviorallooming bias"

Exp. I (EEG)

Fig. 2. Behavioral responses were more consistent for sounds perceived asapproaching compared with sounds perceived as receding if instantaneousspectral changes were presented in continuous stimuli. Mean behavioral re-sponses in the 3–AFC motion discrimination task of Exp. I (Left; n = 15) and the2–AFC motion discrimination task of Exp. II (Middle and Right; n = 13). Resultsof Exp. II are separated between trials presenting instantaneous but continu-ous (Middle; as in Exp. I) and discontinuous (Right) spectral contrast switches.Decreasing spectral contrast switches (orange lines) were predominantly per-ceived as approaching (orange triangles), increasing spectral contrast switches(green lines) as receding (green triangles), and constant spectral contrasts (nolines) as static (gray squares). Statistical analyses focused on these predominantresponse associations. Behavioral looming bias was observed in terms of sig-nificantly higher consistency in approaching responses compared with recedingresponses only if spectral switches occurred within a continuous stimulus.Feedback was provided only on the detection accuracy of constant stimuli (C1 =C2) after blocks in Exp. I. Values reflect mean ±1 SEM. Levels of significance:*P < 0.05, **P < 0.01, ***P < 0.001.

Baumgartner et al. PNAS | September 5, 2017 | vol. 114 | no. 36 | 9745

PSYC

HOLO

GICALAND

COGNITIVESC

IENCE

S

Dow

nloa

ded

by g

uest

on

Nov

embe

r 8,

202

0

pairs, but their magnitudes tended to increase with increases incontrast change. Because the scalp topography of the switch-N1showed a right-hemispheric dominance, we confirmed that switch-N1 amplitudes evaluated at the more frontolateral electrode siteF4 yield similar statistical results (contrast pairs: F1.7,24 = 7.2,P = 0.005, η2p = 0.34; switch direction: n.s.).In contrast to switch-N1, switch-P2 amplitudes differed signifi-

cantly, not only across contrast pairs (F1.7,24 = 16, P < 0.001, η2p = 0.54)but also switch direction (F1,14 = 12, P < 0.01, η2p = 0.46) [in-teraction not significant (F1.6,23 = 0.96, P = 0.38, η2p = 0.064)].Switch-P2 amplitudes for the most salient contrast pair (0 ↔ 1)were significantly larger than for both intermediate pairs (0 ↔ 0.5and 0.5 ↔ 1). Regarding switch direction, switch-P2 amplitudes werelarger for decreasing contrasts (predominantly perceived as approach-ing) than for increasing contrasts (predominantly perceived asreceding). In other words, the amplitude of the switch-P2 was

greater for approaching sounds and thus constituted a neuralcorrelate of auditory looming bias.A post hoc cluster-based permutation test was performed to

quantify statistical differences in the scalp distribution and thetiming of the elicited neural responses for increasing vs. de-creasing spectral contrast. This test revealed that switch directionwas best represented by a single spatiotemporal cluster with abroad central topography and latencies between 120 ms and200 ms. Fig. 3C shows the spatiotemporal coverage on top of atopographic representation of potential differences betweenswitch directions averaged across listeners and contrast pairs.Potential differences of up to 1 μV occur around the centralelectrode Cz and around latencies of 160 ms in favor of switchdirections with decreasing spectral contrasts (approachingsounds). This potential difference was ≈1 μV in magnitude, orabout half the size of the maximum grand-average peak ampli-tude (switch-P2 at Cz).

DiscussionWe investigated the mechanisms underlying auditory looming biaswith combined EEG and psychoacoustic measurements in an in-dividualized virtual auditory environment. Our results show thatlooming bias can be elicited by changes of spectral cues, whilemaintaining overall sound intensity. Interestingly, the bias onlyarises if the stimulus is temporally continuous, leading to perceivedmotion; it is not present when there is a temporal gap betweenpresentations of sounds at different perceived distances. Loomingbias was observed both in increased response consistency and in-creased central cortical activity in the early P2 time range.

Spectral Contrast Manipulation Affected Auditory Distance Perception.Consistent with previous studies on distance perception andsound externalization (27–29), our individualized spectral cuemanipulations affected perceived distance. Natural spectral cues(C = 1, measured at a distance of 1.5 m) created externalizedauditory percepts that were perceived as farthest from the lis-tener, whereas flattened spectra (C = 0) created percepts ofsounds that were very close, often “internal” to the head (29). Ineveryday life, internalized sources are often perceived when lis-tening via head- or earphones, where natural spectral cues arenot present. The intermediate spectral contrast (C = 0.5) in ourexperiment is not necessarily physically plausible, but it showsthat perceived distance gradually decreases with reduced salienceof spectral cues.A recent psychoacoustic study showed gradual degradations of

elevation localization based on very similar spectral manipula-tions (33), which suggests that spectral cues affect localizationperformance similarly in both elevation and distance. In fact,some of our listeners associated spectral cue manipulations atsome source angles with changes in elevation; this is why weselected the source angles used in subsequent testing separatelyfor every individual. The presence of a looming bias in our motiondiscrimination task confirms that perceptual changes mainly oc-curred in the intended distance dimension and not in elevation.The nature of cues that lead to changes in distance perception

can only be debated. Initially we intended to manipulate thesalience of monaural spectral shape cues that are arguably mostimportant for sound localization within a sagittal plane (34).Monaural loudness predictions, however, revealed that smallerspectral contrasts were accompanied by a decrease in loudness atlow and mid frequencies and an increase in loudness at highfrequencies. Hence, systematic changes of sound intensity at highfrequencies could help explain the apparent looming bias. Forlateral source directions, the frequency-dependent loudnesschanges also differed between ipsi- and contralateral ears suchthat low-frequency ILDs systematically increased for smallerspectral contrasts. This frequency-dependent ILD change is in thesame direction as naturally occurring, distance-dependent changes

Fig. 3. Event-related potentials (n = 15) evoked by stimulus onset (A) andstimulus switch (B and C). (A and B) Event-related potential waveforms(A and B, Left) and extracted N1 and P2 amplitudes (A and B, Middle) atvertex electrode Cz. Note that central P2 activity reflects distance perceptionat stimulus onset and looming bias at stimulus switch. Error bars reflect SEM.Levels of significance: *P < 0.05, **P < 0.01, ***P < 0.001. (A and B, Right)Grand average topographies for N1 and P2 latency windows. (C) Significantcluster in time and space (sites marked by asterisks) that is distinctive be-tween decreasing and increasing spectral contrasts overlaid on the corre-sponding difference topography.

9746 | www.pnas.org/cgi/doi/10.1073/pnas.1703247114 Baumgartner et al.

Dow

nloa

ded

by g

uest

on

Nov

embe

r 8,

202

0

of ILDs (35), which is arguably a prominent auditory distance cue(18, 29, 36). Further experiments and/or modeling analyses will berequired to clarify the perceptual weighting of these potentialdistance cues.

Perception of Motion in Distance but Not Changes in Overall LoudnessAre Required to Induce Looming Bias. As summarized earlier,looming bias was previously represented in a broad variety ofpsychoacoustic measures that are either only remotely (e.g.,ratings of loudness change or duration) or indirectly (e.g., ab-solute distance estimation) related to motion perception in dis-tance. Instead of using more inferential measures, here, wedirectly measured response consistency in a forced-choice mo-tion discrimination task because it was not necessarily clear thatperceived distance, and thus induced looming biases, would beconsistently affected using our spectral manipulation technique.Our analysis shows that there was a small concomitant change

in predicted overall loudness that opposed the dominant perceptof motion direction. Thus, our data demonstrate that changes inoverall sound intensity are not required to elicit looming bias.Moreover, in previous looming studies, overall sound intensitywas usually changed slowly over time to simulate motion indistance. Here, we elicited looming bias by a near instantaneous(cross-fade of 10 ms) change of spectral cues in an ongoingstimulus. If the stimulus was interrupted by a short gap (100 ms)during which spectral cues changed, looming bias did not occur;these stimuli were likely perceived as two different auditoryobjects at different distances, rather than a single auditory objectapproaching or receding the listener. Distance perception itselfwas, however, negligibly affected by stimulus dis/continuity asindicated by similar overall response consistencies.Previous studies also argued that noise stimuli are less or not at

all effective in inducing looming bias compared with tone com-plexes, because noise stimuli might be less likely perceived as asingle auditory object (2, 11). The fact that we found looming biasbased on noise stimuli suggests that this stimulus dependencemight be specific to intensity-based distance simulations.Taken together, our behavioral findings suggest that looming

bias does not require specific tasks, distance cues, or stimuli forits activation; it rather constitutes a general increase of saliencewhenever sounds are perceived as moving toward vs. away froma listener.

Cortical Activity After 120 ms Reflected Salience of Spectral Cues andPerceptual Looming Bias. At stimulus onset, the similarity of onset-N1 amplitudes across conditions underscores that sound intensi-ties were similar across spectral contrast conditions (37). Unlikeonset-N1, central onset-P2 amplitudes increased with spectralcontrast, i.e., salience of spectral cues. For sound localization,spectral cues are more important in the median plane comparedwith the horizontal plane. Consistent with our findings, a previousmagnetoencephalography (MEG) study found that spatial devi-ants in the median plane were associated with increased neuralresponse amplitudes within a similar latency range (200–250 ms)as our onset-P2, whereas spatial deviants in the frontal horizontalplane elicited earlier differences (100–150 ms) (38). Moreover, aprevious ERP study tested spectral changes comparable to ours ina left- vs. rightward motion discrimination task and did not findany differences in onset-P2 amplitudes (25). For this task, how-ever, spectral cue evaluation was not essential, whereas in ourexperiment, listeners had to evaluate changes in distance based onspectral cues. This suggests that in our study, the onset-P2 reflectsearly distance-evaluation processes based on spectral cues thatmay lead the percept of sound source distance.At the onset of the spectral switch, larger spectral changes

elicited larger switch-N1 magnitudes, independent of motion di-rection. Conversely, the switch-P2 magnitude depended on motiondirection and reflected auditory looming bias. These results are in

line with the current functional understanding of how vertical andhorizontal motion is reflected in ERPs. Specifically, switch-N1responses are associated with the process of spectral change de-tection and switch-P2 responses with motion direction evaluation(23–26). Hence, our ERP results suggest a generalization of themotion onset response to the distance dimension and furthersupport the conclusion that motion perception is required to elicitauditory looming bias.Regarding the timing of looming perception, it has been shown

that harmonic sounds increasing in intensity lead to enlarged visualcortical activity as early as 80 ms from stimulus onset (15, 22). Ourresults find looming bias reflected after 120 ms from the switch,primarily at central sites associated with auditory processing, whichsuggests that activation of looming bias requires higher-orderprocessing of auditory motion. This apparent discrepancy mayreflect the fact that processing of increments in intensity is lessdemanding and faster than sound motion inferences based onspectral analysis. Another possibility is that, at early processingstages, sound intensity increments are not associated with changesin distance but are simply salient events. For instance, interauralspatial cues occurring at intensity increments are weighted stronglywhen computing left–right location (39, 40). A strong effect of left–right congruence between increasing-intensity sounds and visualtargets (41) further supports this reasoning.Listeners showed large interindividual differences in the amount

of behavioral looming bias they exhibited. Preliminary analysesshowed that switch-P2 amplitudes and measures of global fieldpower might be able to explain these interindividual differenceswith a potential specificity in gender (see Correlation Between In-terindividual Neural and Behavioral Looming Biases and Fig. S5).While our sample size was sufficient to manifest the strong effectof looming bias within subjects (η2p ≥ 0.33), future investigationswith larger and gender-balanced sample sizes are needed to explorethe marked interindividual differences in the amount of behaviorallooming bias.In summary, our results demonstrate that distance cues other

than overall sound intensity can elicit auditory looming bias andthat its activation requires higher-level auditory processing relatedto motion perception. Previous studies on auditory looming biaswere exclusively based on sound intensity as a distance cue andmight have overinterpreted the effect of this particular stimulusfeature. Future studies should review the cortical origins of au-ditory looming bias based on different modes of activation.

Materials and MethodsSubjects. We tested 15 paid volunteers within the age of 20–29 y (M = 24,SD = 3.7; 10 females, 5 males) in Exp. I and 12 paid volunteers plus one co-author within the age of 20–42 y (M = 29, SD = 5.7; 7 females, 6 males) in Exp.II. A subset of five listeners participated in both experiments. None of thelisteners had any indication of a hearing loss greater than 20 dB relative to thethreshold of the normal-hearing population, as confirmed by audiometricthresholds for frequencies from 500 Hz to 8 kHz. All subjects gave informedconsent as overseen by the Boston University Institutional Review Board.

Stimuli and Procedure. Sound source locations were simulated based on in-dividualized measurements of head-related transfer functions (HRTFs)(measurement procedure described inMeasurements of Head-Related TransferFunctions). Stimuli consisted of two consecutively presented Gaussian whitenoises filtered by the band-limited HRTFs with a specific spectral contrastfactor, C∈ f0, 0.51g. The magnitude spectrum, M1ðfÞ, in decibels of the mea-sured HRTFs was manipulated within a frequency range between 1 and 16 kHz(Nf frequency bins) according to C as:

McðfÞ=CM1ðfÞ+ ð1−CÞ 1Nf

X

k∈f

w’ðkÞ M1ðkÞ,

with a frequency weighting function, w ’, that approximates auditory fre-quency resolution by the across-frequency derivative of equivalentrectangular bandwidths (ERBs) (42). The measured phase spectrum remainedunmodified. Filtered noise stimuli were band limited between 1 kHz and16 kHz by a fourth-order Butterworth filter. Two filtered noise representations

Baumgartner et al. PNAS | September 5, 2017 | vol. 114 | no. 36 | 9747

PSYC

HOLO

GICALAND

COGNITIVESC

IENCE

S

Dow

nloa

ded

by g

uest

on

Nov

embe

r 8,

202

0

were cross-faded by 10-ms sine ramps to form a stimulus pair. This stimuluspair was faded in and out by 50-ms squared sine ramps. Details on stimulusdelivery are provided in Stimulus Presentation and Experimental Procedure.

In Exp. I, 840 trials were presented per listener. The spectral contrastswitched in 86% of all trials (14% constant trials). In Exp. II, 192 trials werepresented and spectral contrast always switched (no constant trials). Thepresentation order of trials was completely randomized in both experiments.Listeners received block-based feedback only on their detection accuracy ofconstant trials in Exp. I (no feedback in Exp. II). Feedback was never providedon switch trials to prevent listeners from learning to use cues other thanperceived distance (e.g., timbre) for discrimination. Further details on theexperimental procedure are also provided in Stimulus Presentation andExperimental Procedure.

EEG Recordings. The EEG of 32 scalp electrodes (Activetwo system withActiveview acquisition software, Biosemi B.V.) was recorded in Exp. I withstandard 10/20 montage. In addition, one vertical and two lateral eyeelectrodes were recorded to better capture eye blinks and saccades, andexternal reference electrodes were placed at the mastoids (not used) and earlobes. Real-time processing hardware (RP2.1, Tucker Davis Technologies, Inc.)marked timing of critical experimental events, which were recorded to an

additional data channel alongside the EEG data. The postprocessing pro-cedure for EEG data analysis is described in Processing of EEG Data.

Statistics. For statistical analyses, we used the Matlab Statistics Toolbox. Weconducted repeated-measures analyses of variance with Greenhouse–Geissercorrection (functions: fitrm, ranova). Tukey’s honest significant differencetests were conducted for post hoc paired comparisons (multcompare). Ef-fects were considered significant at P < 0.05.

Data Availability. Code for stimulus creation (sig_baumgartner2017looming),HRTFs, and experimental results (both data_baumgartner2017looming) areintegrated in the Auditory Modeling Toolbox (43). EEG recordings andanalysis scripts are provided via Zenodo (44).

ACKNOWLEDGMENTS. We thank Gerald Kidd Jr. and Norbert Kopco forfruitful discussions and Tim Streeter and Leonard Varghese for technicalsupport. This work was mainly funded by R.B.’s individual fellowship fromthe Austrian Science Fund (Project SpExCue, Grant J3803-N30). Additionalfunding was provided by the European Commission (Project ALT, Grant 691229)and the Boston University Hearing Research Center.

1. Riskind JH, Kleiman EM, Seifritz E, Neuhoff J (2014) Influence of anxiety, depressionand looming cognitive style on auditory looming perception. J Anxiety Disord 28:45–50.

2. Neuhoff JG (1998) Perceptual bias for rising tones. Nature 395:123–124.3. Bach DR, Neuhoff JG, Perrig W, Seifritz E (2009) Looming sounds as warning signals:

The function of motion cues. Int J Psychophysiol 74:28–33.4. Seifritz E, et al. (2002) Neural processing of auditory looming in the human brain. Curr

Biol 12:2147–2151.5. Grassi M (2010) Sex difference in subjective duration of looming and receding sounds.

Perception 39:1424–1426.6. Neuhoff JG (2016) Looming sounds are perceived as faster than receding sounds.

Cogn Res Princ Implic 1:15.7. Neuhoff JG, Planisek R, Seifritz E (2009) Adaptive sex differences in auditory motion

perception: Looming sounds are special. J Exp Psychol Hum Percept Perform 35:225–234.

8. Neuhoff JG, Hamilton GR, Gittleson AL, Mejia A (2014) Babies in traffic: Infant vo-calizations and listener sex modulate auditory motion perception. J Exp Psychol HumPercept Perform 40:775–783.

9. Bach DR, et al. (2008) Rising sound intensity: An intrinsic warning cue activating theamygdala. Cereb Cortex 18:145–150.

10. Deneux T, Kempf A, Daret A, Ponsot E, Bathellier B (2016) Temporal asymmetries inauditory coding and perception reflect multi-layered nonlinearities. Nat Commun7:12682.

11. Ghazanfar AA, Neuhoff JG, Logothetis NK (2002) Auditory looming perception inrhesus monkeys. Proc Natl Acad Sci USA 99:15755–15757.

12. Fay RR, Chronopoulos M, Patterson RD (1996) The sound of a sinusoid: Perception andneural representations in the goldfish (Cassius auratus). Aud Neurosci 2:377–392.

13. Teghtsoonian R, Teghtsoonian M, Canévet G (2005) Sweep-induced acceleration inloudness change and the “bias for rising intensities”. Percept Psychophys 67:699–712.

14. Maier JX, Ghazanfar AA (2007) Looming biases in monkey auditory cortex. J Neurosci27:4093–4100.

15. Romei V, Murray MM, Cappe C, Thut G (2009) Preperceptual and stimulus-selectiveenhancement of low-level human visual cortex excitability by sounds. Curr Biol 19:1799–1805.

16. Bach DR, Furl N, Barnes G, Dolan RJ (2015) Sustained magnetic responses in temporalcortex reflect instantaneous significance of approaching and receding sounds. PLoSOne 10:e0134060.

17. McCarthy L, Olsen KN (2017) A “looming bias” in spatial hearing? Effects of acousticintensity and spectrum on categorical sound source localization. Atten PerceptPsychophys 79:352–362.

18. Kopco N, et al. (2012) Neuronal representations of distance in human auditory cortex.Proc Natl Acad Sci USA 109:11019–11024.

19. Warren JD, Zielinski BA, Green GGR, Rauschecker JP, Griffiths TD (2002) Perception ofsound-source motion by the human brain. Neuron 34:139–148.

20. Hall DA, Moore DR (2003) Auditory neuroscience: The salience of looming sounds.Curr Biol 13:R91–R93.

21. Tyll S, et al. (2013) Neural basis of multisensory looming signals. Neuroimage 65:13–22.

22. Cappe C, Thelen A, Romei V, Thut G, Murray MM (2012) Looming signals revealsynergistic principles of multisensory integration. J Neurosci 32:1171–1182.

23. Getzmann S (2011) Auditory motion perception: Onset position and motion directionare encoded in discrete processing stages. Eur J Neurosci 33:1339–1350.

24. Getzmann S, Lewald J (2012) Cortical processing of change in sound location: Smoothmotion versus discontinuous displacement. Brain Res 1466:119–127.

25. Getzmann S, Lewald J (2010) Effects of natural versus artificial spatial cues on elec-trophysiological correlates of auditory motion. Hear Res 259:44–54.

26. Getzmann S, Lewald J (2010) Shared cortical systems for processing of horizontal andvertical sound motion. J Neurophysiol 103:1896–1904.

27. Brimijoin WO, Boyd AW, Akeroyd MA (2013) The contribution of head movement tothe externalization and internalization of sounds. PLoS One 8:e83068.

28. Boyd AW, Whitmer WM, Soraghan JJ, Akeroyd MA (2012) Auditory externalization inhearing-impaired listeners: The effect of pinna cues and number of talkers. J AcoustSoc Am 131:EL268–EL274.

29. Hartmann WM, Wittenberg A (1996) On the externalization of sound images.J Acoust Soc Am 99:3678–3688.

30. Kulkarni A, Colburn HS (1998) Role of spectral detail in sound-source localization.Nature 396:747–749.

31. Canzoneri E, Magosso E, Serino A (2012) Dynamic sounds capture the boundaries ofperipersonal space representation in humans. PLoS One 7:e44306.

32. Moore BCJ, Glasberg BR, Baer T (1997) A model for the prediction of thresholds,loudness, and partial loudness. J Audio Eng Soc 45:224–240.

33. Macpherson EA, Sabin AT (2013) Vertical-plane sound localization with distortedspectral cues. Hear Res 306:76–92.

34. Baumgartner R, Majdak P, Laback B (2014) Modeling sound-source localization insagittal planes for human listeners. J Acoust Soc Am 136:791–802.

35. Brungart DS, Rabinowitz WM (1999) Auditory localization of nearby sources. Head-related transfer functions. J Acoust Soc Am 106:1465–1479.

36. Brungart DS (1999) Auditory localization of nearby sources. III. Stimulus effects.J Acoust Soc Am 106:3589–3602.

37. Rapin I, Schimmel H, Tourk LM, Krasnegor NA, Pollak C (1966) Evoked responses toclicks and tones of varying intensity in waking adults. Electroencephalogr ClinNeurophysiol 21:335–344.

38. Fujiki N, Riederer KAJ, Jousmäki V, Mäkelä JP, Hari R (2002) Human cortical repre-sentation of virtual auditory space: Differences between sound azimuth and eleva-tion. Eur J Neurosci 16:2207–2213.

39. Dietz M, Marquardt T, Salminen NH, McAlpine D (2013) Emphasis of spatial cues inthe temporal fine structure during the rising segments of amplitude-modulatedsounds. Proc Natl Acad Sci USA 110:15151–15156.

40. Nelson BS, Takahashi TT (2010) Spatial hearing in echoic environments: The role ofthe envelope in owls. Neuron 67:643–655.

41. Leo F, Romei V, Freeman E, Ladavas E, Driver J (2011) Looming sounds enhance ori-entation sensitivity for visual stimuli on the same side as such sounds. Exp Brain Res213:193–201.

42. Glasberg BR, Moore BCJ (1990) Derivation of auditory filter shapes from notched-noise data. Hear Res 47:103–138.

43. Søndergaard P, Majdak P (2013) The auditory modeling toolbox. The Technology ofBinaural Listening, ed Blauert J (Springer, Heidelberg), pp 33–56, Available at amtoolbox.sourceforge.net. Accessed August 10, 2017.

44. Baumgartner R (2017) Data from “asymmetries in behavioral and neural responses tospectral cues demonstrate the generality of auditory looming bias.” Zenodo. Avail-able at https://dx.doi.org/10.5281/zenodo.832899. Accessed August 10, 2017.

45. Brainard DH (1997) The Psychophysics Toolbox. Spat Vis 10:433–436.46. Delorme A, Makeig S (2004) EEGLAB: An open source toolbox for analysis of single-

trial EEG dynamics including independent component analysis. J Neurosci Methods134:9–21.

47. Oostenveld R, Fries P, Maris E, Schoffelen J-M (2011) FieldTrip: Open source softwarefor advanced analysis of MEG, EEG, and invasive electrophysiological data. ComputIntell Neurosci 2011:156869.

48. Maris E, Oostenveld R (2007) Nonparametric statistical testing of EEG- and MEG-data.J Neurosci Methods 164:177–190.

9748 | www.pnas.org/cgi/doi/10.1073/pnas.1703247114 Baumgartner et al.

Dow

nloa

ded

by g

uest

on

Nov

embe

r 8,

202

0